Techniques for accessing computer resources (e.g., application programs and computer data) require that the accessing user be authorized to access the computer resource. The authorization is typically verified using a logon procedure in which a user supplies their credentials (e.g., user identifier and password) to the system that controls access to the computer resource. When the system receives the credentials, it determines whether the user is authorized to access the resource (e.g., by checking capabilities associated with the user identifier). If so, then the system allows the user to access the computer resource. For example, a database program may require a user to supply their user identifier and password before accessing (e.g., reading and writing) the data in the database.
Such techniques for accessing computer resources are well adapted to environments in which the computer resources (e.g., data in a database) are provided by a single source. For example, a news service may require a user to logon before accessing any news stories during that logon session. The news service may then bill the user on a monthly basis for all the news stories accessed during the month. Each provider of computer resources typically develops their own logon procedures. A problem with such development is that each provider duplicates the efforts of other providers when developing their own logon procedures. Another problem with such development is that each provider may use different types of logon procedures which tends to confuse users. For example, one provider may require a password to be at least 8 characters while another provider may require that passwords be 5 to 7 characters. Thus, a user who accesses the computer resources of both providers needs to remember different passwords. Even if different providers were to use the same logon procedures, a problem would still exist in that a user would receive separate bills from each provider. This problem may not be significant if a user uses the computer resources of only a few providers, but it is significant when a user (e.g., an Internet user) accesses the computer resources of many different providers (e.g., different news organizations and associations).
These problems could be solved if a procedure was provided by a logon organization that is not related to providers. For example, a user would logon through a logon organization with a single set of credentials and then access the computer resources for different providers. Providers, however, typically find it undesirable to have their logon procedures provided by an unrelated source for several reasons. First, providers prefer to have the displays associated with the logon of their users to be integrated with the look and feel of the user interface provided by the developer. Second, providers may prefer to have provider accounts used when accessing the providers' services and resources. It would be desirable to have a technique that would combine the advantages of the logon organization to avoid these difficulties, and yet enable providers to use established accounts and provider-supplied logon elements.